I tried both with time measurement and it seems that the normal memcpy function is a little bit faster. So whats the point in using cudaMemcpy?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| cudaMemcpyAsync slower than cudaMemcpy? | 1 | 3086 | March 10, 2009 | |
| cudaMemcpy | 0 | 1208 | November 20, 2008 | |
| cudaMemcpyDeviceToHost taking much time? | 3 | 2655 | July 15, 2009 | |
| cudaMemcpy timing | 1 | 6779 | December 8, 2010 | |
| Why is there the difference of memory copy speed between cpu>gpu and gpu>cpu | 3 | 1280 | April 10, 2014 | |
| Can anyone explain the difference in time? | 2 | 2455 | November 21, 2008 | |
| __pipeline_memcpy_async VS cuda::memcpy_async | 0 | 55 | November 26, 2024 | |
| cudaMemcpy - consuming lot of execution time | 0 | 405 | April 16, 2018 | |
| cudaMemcpyToSymbol | 1 | 5495 | March 4, 2008 | |
| Kernel Copy vs. cudaMemcpy | 1 | 5414 | January 19, 2014 |